Manifesto

We believe the future of compute is decentralized, and that the cleanest way to fund it is to make a token mean something real. On p, every token is bonded to a live inference node, real GPUs answering real requests, and together those nodes form a single decentralized inference grid.

Launching is not minting a meme. It is provisioning capacity. Each token brings a node online, and the node joins the grid the moment it launches. The token is a claim on that capacity, and its price is the value of the compute behind it.

On the other side, anyone, human or agent, can rent that compute through a single endpoint and pay per request. Demand routes to whichever node serves it cheapest. Capacity that earns, grows. Capacity that doesn't, goes dark.

We do not curate and we do not promise returns. We bring the capacity online and let the grid price the only thing here that cannot be faked: GPU time doing useful work.